IT IS an old and well-established fact in social psychology that individuals' judgments, when measured in group situations permitting communication, differ from their judgments when measured in alone situations ("alone" commonly meaning an individual in the presence of an experimenter). The difference characteristically takes the form of a "regression" of judgments towards a group mean. A multitude of interpretations are given to explain this finding. The experiments examined by the present authors (4,5,12,15,16,17) all demonstrate judgmental regression, but there is no conclusive evidence in any of them that actual perceptual changes also occur; the possible exception in some of Asch's subjects (4) has been seriously questioned by F. Allport (3, pp. 3671.). When one examines the types of judgments required, one realizes the hazards of generalizing the principle of judgmental regression to social behaviour. In the experiments mentioned above, subjects were asked to judge the lengths of lines, the sizes of rectangles, the number of beans in a bottle, the number of clicks heard, the imagined movement of a spot of light, and the lengths of sticks, respectively.It is not only the "communicating'' feature of the group that leads to judgmental regression, for F. Allport (2) reports that the mere proximity of co-workers, with no communication permitted, tended to moderate judgments of the pleasantness or unpleasantness of odours and of the heaviness of weights.The purpose of the present study is to determine if attitudinal responses, the nature of which constitutes a central problem in social psychology, are susceptible to regression towards the group mean as judgments of external stimuli have been shown to be. Farnsworth (8) asked college students to judge eighteen statements as to their degree of pacifism or militarism, testing different samples of students in the alone and in the group situations, and concluded that there was no marked difference in their judgments. LaPiere and Farnsworth later stated that "no ... tendency toward social conformity appeared in [this] study of authors wish to thank Professors G. A. Ferguson and J. D. Ketchum for their generous assistance in the organization and presentation of this study, and Mr. R. Gardner who kindly checked the statistical findings.
SUMMARYSixty-five male undergraduates filled out comparable forms of the F scale (1) under three conditions: alone, in groups of five without discussion, and in groups of five after discussion. The over-all findings indicate that group discussion produces a significant reduction in variability of attitude scores, but the mere presence of others does not. When attention is given to the degree of acquaintance among Ss, however, a significant reduction in score variability is evident for High Acquaintance Ss from the alone to the together situation, where no communication is permitted. Low Acquaintance Ss appear attitudinally unaffected by either the presence or discussion of others. A hedonistic explanation is offered for the findings.